Research on fault feature extraction for planet-bearing using novel combination operator morphological filtering and stochastic resonance

Abstract

Vibration transmission path of planet-bearing varies with the revolution of the planet gear, resulting in complex sideband interference near the fault feature frequency. Additionally, gear meshing vibration and strong noise interference also increase the difficulty of fault feature extraction for planet-bearing. To address this, this paper introduces a novel combination operator morphological filtering (NCOMF) method to effectively reduce noise and enhance fault impact features. The structural elements of different scales match the local features of planet-bearing signals to obtain the morphological filtering results at multiple scales. Furthermore, this paper utilizes the cloud genetic algorithm to screen out the optimal weighting values of each scale. Teager energy operator statistical complexity measures index could be employed as a fitness function to evaluate the fault impact response of vibration signals. Consequently, the morphological filtering signals at different scales are bound with weights to optimally enhance the fault feature of planet-bearing inner ring. Subsequently, the weighted filtered signal is inputted into the piecewise tri-stable stochastic resonance (SR) system, which leverages noise benefits to enhance the fault vibration signal energy of planet-bearing inner ring. Ultimately, the fault feature of planet-bearing inner ring is successfully extracted using the combination of NCOMF and piecewise tri-stable SR system.